Organic Photoactive Materials with Aggregation-Induced Emission Characteristics for Tumor Theranostics

3.2.1 Subcellular organelle-specific targeted photoactive materials

As the critical challenges associated with the TME, such as heterogeneity, hypoxia, and immunosuppression, become increasingly prominent, the therapeutic strategies involving photoactive materials are undergoing a profound transformation from single-target breakthroughs to multisystemic regulation^{[77–78–79]}. By precisely localizing the substructures of tumor cells, researchers are integrating photothermal/photodynamic effects with the activation of cell death pathways, thereby gradually establishing a spatiotemporally synergistic precision therapy framework^[80]. This evolutionary process encompasses not only the expansion of targeting dimensions (from single-targeting to dual-targeting) but also the in-depth refinement of single-targeting systems, ultimately propelling tumor therapy from localized interventions to the regulation of metabolic networks^[81].

The therapeutic strategies involving photoactive materials for tumor treatment have undergone significant advancements, transitioning from single-target breakthroughs to multisystemic regulation. Early single-targeting research primarily focused on the precise localization of subcellular structures. The TCF series of AIE PS developed by the Gong Ping group (Figure 3A), designed to target specific organelles such as mitochondria, the ER, and lipid droplets^[82]. The multifaceted effects of a single target were revealed through TCF-Mem, which targets the cell membrane. Upon photoactivation, TCF-Mem induces pyroptosis and releases damage-associated molecular patterns (DAMPs) and inflammatory cytokines. This approach couples local treatment with systemic immune modulation for the first time. This discovery not only confirmed the efficacy of single-target approaches but also underscored the complexities introduced by tumor metabolic compensatory mechanisms. Specifically, TCF-Mito, which targets mitochondria and disrupts energy metabolism, may inadvertently facilitate cancer cell survival via lysosomal autophagy pathways. It provides a critical impetus for the development of dual-targeting synergistic strategies. Building upon the elucidation of single-targeting mechanisms, researchers have observed that the metabolic compensation effects in tumor cells frequently lead to the failure of single-target therapies. Consequently, they have embarked on exploring spatiotemporally precise dual-targeting designs to impede cellular escape pathways. This endeavor has propelled the rapid development of dual-targeting strategies, the crux of which lies in synergistic spatial localization and temporal effect superposition to block cellular escape routes^{[86–87–88–89]}. Zhuang et al.^[83] developed an aggregation-induced NIR-II emissive PS, DPITQ, via molecular engineering. Its lipophilic and cationic nature aligns with cancer cell physiology, featuring high lactate anion expression and hyperpolarized mitochondrial membrane potential (MMP), enabling selective accumulation at both plasma membranes and mitochondria (Figure 3B). This dual-targeting design facilitates spatiotemporally precise activation of type I/II ROS pathways under illumination, concurrently inducing pyroptosis and apoptosis to block tumor cell escape. DPITQ NPs achieve a high photothermal conversion efficiency of 57% while maintaining excellent metabolic stability, resulting in superior phototherapeutic outcomes.

The spatial confinement–mechanistic synergy design philosophy has been further extended and innovated in the microneedle patch system developed by Liu et al.^[90] An intelligent combination of mitochondria-targeting AIE molecules and lysosome-targeting aluminum phthalocyanine (AIE-mito-TPP/AlPcSNa₄@MN) achieves a 99% inhibition rate in melanoma models by disrupting energy metabolism (ATP synthesis) and degradation systems (lysosomal membrane permeabilization). This dual-targeting strategy efficiently kills tumor cells at low doses (6 μg) by synergistically disrupting energy metabolism and degradation systems. Furthermore, the PPRK@MTO nanomedicine developed by Liang et al.^[84] achieves highly efficient synergistic therapy through precise dual-organelle targeting (Figure 3C). This system utilizes an ER-targeting peptide, PPRK, to guide the specific accumulation of the nanomedicine in the tumor cell ER, enhancing the ROS generation efficiency of PDT by 3.1-fold. Simultaneously, the mitochondrial-targeting component, MTO, selectively inhibits the mitochondrial calcium uniporter (MCU) channel, blocking calcium influx and disrupting mitochondrial–ER metabolic coupling. This spatially precise regulation results in a 2.3-fold increase in the expression of the ER stress marker GRP78, causing the collapse of the unfolded protein response by inhibiting ATP/GSH synthesis, and ultimately enhancing dendritic cell (DC) maturation by 50.7%. In the 4T1 breast cancer model, a single administration achieves an 85.28% tumor inhibition rate, underscoring the unique advantages of dual-subcellular organelle targeting strategies. These breakthroughs underscore that dual-targeting strategies can generate a “spatiotemporal resonanc” synergistic effect by intervening at key nodes of the metabolic network. However, their complexity has also spurred the in-depth refinement of single-targeting systems.

Notably, the advancement in dual-targeting research has inversely propelled the intelligent upgrading of single-targeting technologies. The BTZPP NPs developed by Huang et al.^[91] have elevated lysosomal targeting precision to a Pearson correlation coefficient of 0.91, enabling a therapeutic-monitoring closed loop through long-term imaging. The MNBS PS designed by Xiong et al.^[92] triggers ferroptosis by specifically targeting lipid droplets, with an enhanced type I ROS yield compared with conventional systems, unveiling a novel correlation between organelle localization and the regulation of cell death modalities. Feng et al.^[93] have transcended the functional limitations of traditional single-targeting approaches with their OTBS-FR-ER self-assembling peptide. By targeting the ER and undergoing ordered nanofiber assembly, this peptide enhances ROS generation efficiency and triggers an ROS/ reactive nitrogen species (RNS) cascade reaction, leading to upregulated GRP78 expression and successfully activating robust type II immunogenic cell death (ICD). The THTTPy-PTSA probe, developed by Wang et al.^[85] (Figure 3D), represents a more significant breakthrough. Initially designed by incorporating PTSA for ER targeting, this probe undergoes photo-oxidative dehydrogenation of its tetrahydropyridine group upon light irradiation, converting it into a pyridine group that subsequently functions as a mitochondrial-targeting moiety. This dynamic targeting mechanism enables the probe to migrate intelligently from the ER to mitochondria, thereby spatiotemporally amplifying oxidative stress and calcium signaling within both organelles. The resulting effects include elevated expression of ER stress markers (p-PERK/p-eIF2α), decreased mitochondrial membrane potential, and ultimately, activation of a robust immune response. Through modular design, these sophisticated single-targeting systems can function independently as therapeutic units or be flexibly assembled into multitargeting platforms, establishing an iterative cycle of “basic optimization → combinatorial innovation → re-optimization.”

Currently, this field has forged a synergistic evolutionary trajectory for targeting strategies: single-targeting research serves as a cornerstone, providing molecular probes and mechanistic blueprints for multitarget synergism. Breakthroughs in dual-targeting approaches, in turn, inversely propel the intelligent and functional evolution of single-targeting systems. Furthermore, novel single-targeting technologies lay the groundwork for more intricate multitargeting designs. In the future, it is essential to transition from “empirical combinations” to “systematic regulation” by leveraging artificial intelligence (AI)-assisted optimization of targeting ligand combinations, dynamic dissection of interorganelle signaling crosstalk, and the development of carriers with enhanced metabolic stability. This transformative shift will usher tumor phototherapy into a new epoch of precision medicine^[94].

3.2.2 Photoactive materials for tumor microenvironment modulation

The TME exhibits dynamic heterogeneity, resembling a complex pathological network. It is characterized by hypoxia, acidity, elevated interstitial fluid pressure, and immunosuppression, which collectively pose significant barriers to effective tumor therapy (Figure 4)^[95,96]. The advent of photoactive materials is revolutionizing the paradigm of tumor treatment by converting the pathological characteristics of the TME, which once posed therapeutic resistance, into strategic points for precise intervention. This is achieved through intelligent molecular-level design and systematic dynamic regulation^[97,98].

Within the hypoxic TME, the oxygen-dependent mechanism of traditional PDT faces fundamental challenges. However, innovative strategies involving nanomaterials have opened up bidirectional pathways to address these challenges. Zhang et al.^[99] engineered TPASIC-PFH@PLGA NPs, which are loaded with liquid perfluorohexane (PFH). Upon exposure to NIR light, these NPs undergo a liquid-to-gas phase transition. This process not only releases stored oxygen to alleviate local hypoxia but also forms dense aggregates through nanostructural reorganization, significantly enhancing the ROS generation efficiency of the PS in its aggregated state. This “self-oxygenation-enhancement” dual-mode strategy transforms PDT’s oxygen requirement from passive dependence to active supply (Figure 5A). For the deeply seated tumor regions experiencing extreme hypoxia, Zhou et al.^[100] have advanced a novel strategy for targeting hypoxic tumor cores through their BT-LRC nanosystem. This system exploits an electron transfer cascade between the BODIPY-tetraphenylethylene (TPE) fluorophore and camptothecin, enabling direct production of type I ROS in HepG2 cells under severe hypoxia (Figure 5B). Concurrently, it releases DNA-damaging therapeutics, establishing an oxygen-independent “type I photodynamic-chemotherapy” synergistic network. This technological advancement from “oxygen supplementation” to “bypassing oxygen dependence” aligns closely at the molecular level with the NIR-BN nanosystem developed by Shi et al.^[101] This material possesses an exceptionally small singlet–triplet energy gap (ΔE_ST = 0.09 eV) and a high molar extinction coefficient (3.3 × 10⁴ M⁻¹ cm⁻¹), ensuring efficient intersystem crossing (ISC). Its twisted molecular configuration induces multimodal structural relaxation upon photoexcitation, endowing the NIR-BN NPs with a photothermal conversion efficiency exceeding 50%. Crucially, intense intramolecular dihedral angle vibrations in the excited state enable the direct generation of highly toxic superoxide anion radicals (·O₂⁻) under 660 nm laser irradiation, thereby bypassing the dependence on molecular oxygen inherent in conventional PDT. This unique molecular mechanism provides an innovative paradigm for oxygen-independent synergistic PTT/PDT of deep-seated tumors. After nanomaterials surmount the hypoxia barrier within the TME, the acidic characteristics of the TME emerge as a new focal point for intervention. The pH of tumor tissues is typically lower than that of normal tissues (pH 6.5 to 7.0 versus 7.4), providing a rationale for the design of pH-responsive photoactive materials^[102]. These materials often incorporate pH-sensitive chemical bonds or groups, including hydrazone bonds, imine bonds, and carboxylic acid groups. These components undergo cleavage or conformational changes in acidic environments, facilitating controlled drug release^{[103–104–105]}. Dual-stimuli-responsive nanocarriers, which integrate pH-sensitive polymers with PSs, exhibit charge conversion under acidic or alkaline conditions through protonation and deprotonation mechanisms. These carriers also amplify the therapeutic efficacy of PTT or PDT upon light irradiation^[106]. In the carrier-free HD-APNNA NPs with AIE characteristics designed by Cheng et al.^[107], this concept is transformed into an intelligent responsive switch: protonation of carboxylic acid groups drives a reversal of surface charge from −5 to +10 mV (Figure 5C), accompanied by dynamic shrinkage of the NP size to 170 nm. This “acid-triggered dual response in morphology and function” not only enhances cellular membrane adsorption and tissue penetration but also achieves an ROS quantum yield of 56.7%, surpassing that of the clinical PS Ce6, while simultaneously avoiding the biological toxicity risks associated with exogenous carriers. This “acid-overcoming-acid” strategy is further expanded in a multidimensional manner in the NAB@DD-DC system developed by Du et al.^[108] Here, pH-sensitive polymers hydrolyze in the acidic lysosomal environment, triggering charge reversal to facilitate lysosomal escape. The released Na⁺ disrupts mitochondrial membrane potential through osmotic pressure, concurrently activating mitochondrial-targeted imaging and pyrimidine metabolism blockade. This approach transforms the acidic microenvironment of the TME from a therapeutic obstacle into a spatiotemporal control node for precise delivery and synergistic therapy (Figure 5D).

The heightened interstitial fluid pressure and dense extracellular matrix (ECM) architecture within the TME have propelled innovative explorations in photoactive materials, employing physical–chemical synergistic strategies to breach barriers^[110]. Using glioma as a representative model^[111], the unique characteristics of its microenvironment further exacerbate therapeutic challenges: the blood–brain barrier (BBB) strictly impedes drug penetration^[112,113], deep-seated infiltrative lesions are difficult to excite PSs due to tissue light attenuation^[114], and hypoxic regions induced by rapid tumor proliferation directly attenuate the oxidative cytotoxic effects of conventional PDT^[115]. To address this multidimensional barrier system, Xu et al.^[109] and Zhang et al.^[116] have proposed groundbreaking solutions. The former achieves reversible BBB opening and precise light-controlled therapy at a depth of 940 μm by combining focused ultrasound with a 2-photon-excitable NIR-II PS, enabling the simultaneous generation of both type I and type II ROS to cope with oxygen gradient heterogeneity. The latter designs an Angiopep-2 receptor-targeted type I PS and utilizes an air-core modified optical fiber to deliver laser light in situ to the tumor core, efficiently generating hydroxyl radicals (·OH) and superoxide anions (·O₂⁻) within a depth of 25 mm, thereby surmounting the limitations imposed by the hypoxic microenvironment (Figure 5E). The 2 strategies, centered on “physical modulation-deep energy delivery” and “biological navigation-in situ excitation,” systematically dismantle the “barrier-hypoxia-light attenuation” dilemma in the glioma microenvironment, providing clinically translatable technical paradigms that combine precision and versatility. Within the technological pursuit of targeting the TME, photoactive materials are evolving from passive adaptation strategies to proactive architectural designs^[117]. These advanced materials possess the capability to monitor multifaceted physiological variables, including hypoxia, acidity, and mechanical stress in real time, while concurrently converting pathological traits into therapeutic impetus through intricate molecular engineering. Specifically, phase-change substances restore oxygen homeostasis, protonation-responsive mechanisms bypass delivery impediments, electron transport systems override oxygen limitations, and mechanochemical interactions redefine immune dynamics^{[118–119–120–121–122]}. Looking ahead, these dynamic regulatory systems may integrate optogenetic elements to achieve synergistic gene–microenvironment editing. Additionally, AI-driven closed-loop feedback mechanisms could optimize therapeutic parameters in real time. Furthermore, these systems might construct “energy hijacking” networks by exploiting tumor metabolic reprogramming. The evolution of nanomaterials from mere microenvironment decoders to proactive microenvironment architects signifies a transformative leap in tumor therapy, transcending conventional boundaries of localized cytotoxic effects and embarking on a novel epoch of systemic modulation. This paradigm shift not only necessitates a profound interdisciplinary dialogue between materials science and synthetic biology but also demands a precise reconstruction of the dynamic interplay within the TME using biomimetic models such as organoids and organ-on-a-chip platforms. Ultimately, this will facilitate the elevation of therapeutic strategies from molecular interventions to systemic remodeling, marking a significant advancement in cancer treatment.

3.2.3 Wavelength-tunable photoactive materials

In the field of precision tumor diagnosis and therapy, the evolution of photoactive materials is undergoing a paradigm shift from functional hybridization approaches to intelligent integration systems. The traditional “All-in-One” strategy, which combines diagnostic and therapeutic platforms (e.g., fluorescent dyes and PSs) through physical or chemical hybridization of multiple components, has facilitated multifunctional platform development. However, its clinical translation is severely impeded by challenges such as poor batch reproducibility, with a coefficient of variation frequently surpassing 15%, and unpredictable pharmacokinetics resulting from complex intercomponent interactions^{[123–124–125]}. This predicament has spurred the advancement of novel phototheranostic systems leveraging the AIE mechanism. The core of these systems lies in establishing a delicate balance between radiative and nonradiative decay pathways through intramolecular energy regulation, thereby integrating imaging and therapeutic functionalities within a single molecular species^[126]. Compared with traditional materials, AIE materials not only streamline the preparation process by 3- to 5-fold but also exhibit well-defined structural characterization and optimized biocompatibility, demonstrating disruptive potential for clinical applications^{[127–128–129]}. This innovative paradigm successfully overcomes the aggregation-caused quenching (ACQ) effect observed in traditional dyes. Leveraging the low tissue scattering, weak absorption, and high penetration depth characteristics of NIR light (700–1700 nm), AIE materials provide a technological foundation for the diagnosis and treatment of deep-seated tumors^[130,131]. The technological evolution of NIR-I (700–900 nm) AIE probes follows a paradigm shift trajectory of “functional expansion—mechanistic innovation—technological leapfrogging,” progressively advancing from foundational imaging to ultrahigh-resolution dynamic observation^[132].

Early-stage breakthroughs targeted metabolic limitations of conventional probes. The DPPM-TPA NPs developed by Han et al.^[133] employ a donor–acceptor–donor symmetric architecture (diphenylamino donor/DPPM core acceptor), achieving NIR absorption at 475 nm and emission at 585 nm. By leveraging 2-photon excitation technology, these NPs extended the in vivo tumor fluorescence tracking period to 9 days (Figure 6A), establishing a foundation for long-term imaging. Subsequent research prioritized theranostic integration, as exemplified by the TTCBTA NPs pioneered by Ma et al.^[134]. These NPs are based on a donor–acceptor structural design (triphenylamine as the electron donor and benzothiazole as the electron acceptor), with molecules exhibiting highly twisted conformations. This spatial steric hindrance effectively suppresses π–π stacking interactions, reducing nonradiative decay pathways in aggregated states and thereby overcoming ACQ phenomena. Additionally, the strong intramolecular charge-transfer characteristics induce significant structural relaxation and solvent reorganization upon excitation, resulting in a substantial reduction in excited-state energy and a large Stokes shift of up to 230 nm. This significant spectral separation effectively reduces interference from fluorescence self-absorption and spontaneous fluorescence from biological tissues, enhancing signal penetration capability and facilitating high-contrast imaging of deep tissues. Concurrently, Yan et al.^[135] designed a water-soluble AIE material, PEGTPA-NSD with enhanced biocompatibility, facilitating prompt signal accumulation in models of breast cancer bone metastasis. The structural features of PEGTPA-NSD indicated its potential for extended emission into the NIR-II region, motivating preliminary investigations into NIR-II imaging applications (Figure 6B). Building on functional integration, Barman et al.^[136] advanced a luminescent mechanism innovation by integrating AIE characteristics with thermally activated delayed fluorescence, engineering BTMCz NPs. These NPs significantly improved PS imaging contrast and therapeutic efficacy at the photophysical level, offering theoretical support for super-resolution advancements. Ultimately, Xu et al.^[137] successfully enhanced the solid-state photoluminescence quantum yield of their cationic AIEgen, DTPAP-P, to 35.04% by implementing a dual-pathway strategy that suppresses molecular motion. The 700 nm emission peak, combined with stimulated emission depletion nanoscopy, enhanced mitochondrial imaging resolution from 1028 to 165 nm. It also introduced a light-controllable targeting switch mechanism, enabling dynamic tracking of mitochondrial fission–fusion in live cells at 165 nm resolution and nucleolar migration for super-resolution imaging in fixed cells at 184 nm (Figure 6C). This advancement addresses clinical needs for “prolonged visualization and precise intervention” while revealing, through spatiotemporally controlled nanoscale observation, molecular-scale interactions between mitochondria and nuclei for the first time. Consequently, this represents a transformative shift of AIE probes from theranostic tools to decoders of organelle interactions. However, NIR-I imaging is limited by inadequate tissue penetration depth (<3 mm) and significant signal attenuation (>50%) in deep tumors. Furthermore, challenges such as photobleaching and background noise interference hinder the accurate detection of small lesions (<2 mm)^[138]. Consequently, research has increasingly focused on the second NIR-II (1000–1700 nm), which offers superior penetration capabilities. The photon scattering coefficient in the NIR-II region is reduced by 2 to 3 orders of magnitude compared with NIR-I, enabling penetration depths exceeding 5 mm. Additionally, autofluorescence from biological tissues is significantly diminished in this range, substantially improving the signal-to-background ratio (SBR) in imaging^[139].

In recent years, multimodal theranostic systems based on NIR-II AIE materials have achieved significant advancements due to breakthroughs in molecular engineering. Research now focuses on the precise modulation of emission wavelengths and the synergistic enhancement of diagnostic and therapeutic functionalities^[140,141]. From the perspective of the technological evolution trajectory, where emission wavelengths progressively shift from shorter to longer regions, various research teams have advanced through innovative molecular design strategies, systematically overcoming bottlenecks in deep-tissue imaging and synergistic therapy. A landmark achievement in the short-wavelength NIR-II region (900–950 nm) was reported by Jin et al.^[142], who developed SFX-IC NPs. They employed a spirocyclic functionalization strategy to develop organic PSs with an emission peak precisely positioned at 935 nm when excited by an 808 nm laser. These PSs exhibited a fluorescence brightness 1.74-fold higher than that of the commercially available ICG. This molecular engineering not only achieved fluorescence imaging with a high SBR of >4.2, significantly enhancing imaging clarity and accuracy, but also integrated multimodal imaging capabilities, including PAI and PTI, enabling multidimensional and complementary biomedical monitoring. Moreover, the rigid spirocyclic structure effectively suppressed nonradiative transitions, providing a new paradigm for the precise localization of deep-seated tumors (e.g., subcutaneous lesions at 7 mm). This achievement laid an important foundation for the subsequent development of longer-wavelength probes. Next, Zheng et al.^[143] successfully extended the emission wavelength to 950 nm through a synergistic approach in receptor engineering and sulfur atom embedding technology. The designed 4TPQ NPs demonstrated excellent fluorescence properties while strategically leveraging the heavy-atom effect of sulfur atoms and the charge-transfer characteristics of the receptor units. These design elements enhanced ROS generation efficiency and achieved a photothermal conversion efficiency of 32.5%. This dual-functionalized “imaging-therapy” design exhibited unique advantages in a mouse model of breast cancer bone metastasis. NIR-II fluorescence guidance accurately localized micrometastases, while the synergistic photothermal and PDT reduced tumor volume, forming a complete technological chain from molecular design to clinical translation. Along the technological route of wavelength extension, Zhang et al.^[144] further pushed the emission wavelength to 992 nm through an intramolecular motion regulation strategy, constructing the breakthrough TSSI NP system. This material exhibits multiple advantages in the far-end NIR-II region (>950 nm): first, its NIR-II fluorescence imaging SBR is enhanced, enabling the detection of minute tumors (<2 mm) that are difficult to identify with traditional methods; second, the material’s unique dual-channel energy conversion mechanism allows for high-resolution imaging of tumor microvascular networks through PAI and synergistic tumor-killing effects through PTT and PDT under NIR excitation. Notably, the PTI functional module enables real-time monitoring of the temperature field distribution in the treatment area, providing a dynamic feedback mechanism for precise control of photothermal conversion efficiency (Figure 7A). In the 4T1 breast cancer model, this multimodal synergistic mechanism achieved a dynamic closed loop for real-time monitoring of therapeutic responses and efficacy evaluation, marking a new stage in the advancement of “theranostic integration” technology. These 3 typical cases present a clear evolutionary logic: from the short-wavelength high-brightness imaging of SFX-IC to the treatment-enhancing mid-wavelength 4TPQ, and then to the long-wavelength multimodal fusion of TSSI. The researchers have systematically addressed the challenges of balancing tissue penetration depth, therapeutic efficiency, and imaging accuracy through iterative innovations in molecular engineering. This design philosophy of wavelength-function synergistic optimization provides a crucial paradigmatic reference for the future development of intelligent theranostic platforms.

Building on the aforementioned research, You et al.^[145] developed the 4TT-PBPT AIEgen utilizing N-heterocyclic fused-ring design, pushing the emission wavelength to 1026 nm. This material not only enables millimeter-level localization of in situ bladder tumors through trimodal imaging (NIR-II FLI/PAI/PTI) but also achieves a photothermal conversion efficiency of 73.8% at 100 μM, approaching the theoretical limit of gold NPs. Combined with the ROS-mediated apoptosis mechanism, it attains a tumor ablation rate exceeding 95% (Figure 7B, C). This innovation in the fused-ring system significantly enhances material stability and functional density, opening up new avenues for the design of ultralong-wavelength probes. As the emission wavelength extends further into longer spectral regions, the BETT-2 AIE molecule developed by Song et al.^[146] exhibits an emission peak at 1240 nm. This material efficiently generates superoxide radicals (O₂⁻) through the type I photodynamic pathway, while simultaneously achieving a photothermal efficiency of 56.6%. Its multimodal imaging system enables dynamic analysis of TME heterogeneity, providing novel insights for addressing the challenges of energy delivery to deep-seated tumors. More cutting-edge is the “inside-out” phototherapy strategy proposed by Zhang et al.^[147], which integrates NIR-II AIEgens with modified optical fibers (Figure 7D). By introducing heavy atoms and strong donor–acceptor interactions and highly twisted conformations, the molecule exhibits NIR-II fluorescence emission (1092 nm), efficient photothermal conversion (η = 58.9%), and photodynamic effects, and photodynamic effects primarily generating hydroxyl and superoxide radicals under 808 nm excitation. NPs surface-modified with the iRGD peptide (NPs-iRGD) are capable of targeted enrichment at tumor sites, and modified optical fibers, comprising a fused flat-end fiber and an air-core fiber, are utilized to deliver light energy to deep tissues. Furthermore, this strategy maintains efficient photothermal/photodynamic effects in skin barrier simulation experiments, demonstrating its potential for clinical translation.

In recent years, photoactive materials excited in the second NIR-II have revolutionized tumor vascular imaging technology, leveraging their exceptional deep-tissue penetration and precise optical tomography capabilities. Through synergistic innovations in molecular engineering and nonlinear optics, researchers have constructed multimodal imaging probe systems, achieving stepwise breakthroughs in imaging depth, spatiotemporal resolution, and functional visualization. The pioneering work in this field was conducted by Li et al.^[148], who developed the NIR-II AIE luminogenic DCTBT. By precisely tuning the electron-donor structure, this probe achieved unprecedented brain imaging depths of 2180 μm (with skull removed) and 1135 μm (intact skull) under 1700 nm 2-photon excitation. Boasting a large 2-photon action cross-section of 1.2 GM and retaining 90% of fluorescence, it not only sets new records for penetration depth among organic probes but also facilitates the effective monitoring of cerebrovascular hemodynamics at a depth of 714 μm, laying a technological foundation for studying deep-seated tumor vasculature (Figure 8A). This breakthrough has spurred interdisciplinary investigations into multidimensional imaging, notably culminating in the seminal work of Gong et al.^[149], who achieved a pivotal advancement in 3-dimensional (3D) quantitative analysis. They developed HA@TANP NPs, leveraging synergistic temperature-controlled polymerization and optical clearing techniques to facilitate 3D panoramic reconstruction of the vascular network in metastatic lung cancer models. Achieving a spatial resolution surpassing the 5 μm limit, this system allows precise quantification of dynamic changes in vascular diameter (4–55 μm), length (25–75 μm), and morphological straightness (0.899–0.914). This advancement has propelled the study of tumor vascular heterogeneity from 2D observational paradigms into a new era of 3D quantitative analysis.

With the advancement of 2PF techniques, 3PMhas emerged as a transformative tool. It capitalizes on the enhanced deep-tissue penetration capabilities of the NIR-III excitation window, effectively addressing the depth limitations inherent in 2PF for in vivo brain imaging. Early research efforts have centered on the development of noninvasive probes. For example, Qin et al.^[151] designed the AIE molecule—BTF, devoid of a traditional TPE backbone, which exhibits a large 3-photon absorption cross-section of 2.56 × 10⁻⁷⁹ cm⁶ s² under 1550 nm excitation. High-quantum-yield (36.1%) probes were prepared using a nanoprecipitation method. This groundbreaking work enabled 3D imaging of the cerebrovascular network beneath intact skulls (penetration depth: 900 μm), facilitating the first noninvasive monitoring of cerebral thrombosis formation and providing an in situ observational tool for cerebrovascular pathology research. Advancements in molecular engineering have propelled the performance limits of 3-photon fluorescence probes to new heights. Yang et al.^[150] optimized an AIEgen—TIT, featuring a symmetric donor–acceptor–donor configuration that precisely couples planar electron acceptors with twisted donor units. Under 1665 nm excitation, TIT achieves a 3-photon action cross-section (ησ₃) of 603 × 10⁻⁸⁴ cm⁶ s² photon^-2. In cranial window models, the probe achieves an imaging depth of 1895 μm and enables dynamic tracking of hemodynamics at depths up to 1200 μm (Figure 8B, C). Compared with the earlier BTF system, TIT enhances NIR-III excitation efficiency by an order of magnitude through a delicate balance of π-conjugation extension and molecular motion suppression, marking a paradigm shift from “structural observation” to “functional analysis” in deep-brain imaging.

The evolution from depth breakthroughs in 2-photon microscopy to multidimensional analysis enabled by 3-photon microscopy represents a significant technological trajectory, characterized by the deep integration of NIR materials and nonlinear optics. Current research not only uncovers in situ characteristics of aberrant vascular branching and leakage within TMEs through advanced optical tomography capabilities but also paves the way for theranostic integration via modular probe design. By integrating targeting recognition and therapeutic functional units, these intelligent nanoprobes hold promise for realizing closed-loop diagnostics and therapeutics—spanning from vascular anomaly detection to precision drug delivery. This development propels the field of precision oncology into a new era of molecular dynamics manipulation, heralding transformative possibilities for personalized cancer treatment.

3.2.4 Photoactive materials for combined immunotherapy strategies

Immunotherapy constitutes a revolutionary approach in oncology, achieving antitumor effects by activating the body’s immune system^[152]. Among these approaches, immune checkpoint blockade therapy stands out for its mechanism of interrupting immunosuppressive pathways such as PD-1/PD-L1. This approach markedly augments the antitumor activity of T cells and has demonstrated remarkable efficacy in treating malignancies like melanoma and non–small cell lung cancer^{[153–154–155]}. However, the systemic administration of traditional monoclonal antibodies frequently triggers immune-related adverse reactions and struggles to overcome the immunosuppressive characteristics of the TME. This has prompted researchers to explore more targeted and synergistic combination therapy strategies^{[156–157–158]}. Phototherapy-induced ICD offers a novel approach to circumvent these limitations^[159]. By triggering the release of danger signals such as calreticulin (CRT), high mobility group box 1 protein (HMGB1), and ATP from tumor cells through PDT or PTT, DCs can be activated and antigen presentation enhanced, thereby strengthening systemic antitumor immune responses^{[160–161–162–163]}.

The technological evolution of photoimmunotherapy has unfolded through a series of progressive breakthroughs, structured across 4 interconnected dimensions: local ablation, systemic activation, precise modulation, and multimodal synergy. During the initial mechanistic investigation, Liu et al.^[164] developed the BSA/TPA-Erdn nanosystem, capitalizing on their distinctive photophysical characteristics. They validated the release of danger signals such as CRT and HMGB1 mediated by ROS via utilizing the NSG mouse models. This confirms the primary immune activation effects of PDT-induced DC maturation and T-cell infiltration (Figure 9A). This study established the molecular design foundations of photoimmunotherapy. However, constrained by its unimodal therapeutic mechanism, the induced ICD effect had not yet breached the physical barriers of the TME. The TPETTBI molecule proposed by Yang et al.^[165] marked a pivotal technological breakthrough. Through molecular engineering, they integrated mitochondrial-targeting modules with AIE-active groups, precisely localizing therapeutic targets to cellular energy hubs. This design achieved spatiotemporal synergy between efficient ROS generation (PDT) and NIR photothermal conversion (PTT). Experimental data demonstrated that this dual-strike strategy, targeting the mitochondrial microenvironment, amplified ICD effects 3.2-fold and successfully activated CD8⁺ memory T-cell subsets, achieving a 67% inhibition rate for distant metastases in bilateral tumor-bearing models (Figure 9B–D). This multimodal synergistic mechanism addressed the issue of limited tissue penetration inherent in conventional phototherapies. Nonetheless, the study simultaneously revealed the limitations of ICD induction alone; although it effectively provokes immediate immune reactions, it struggles to reverse deep-seated immunosuppressive microenvironments characterized by tumor-associated macrophage polarization and regulatory T-cell infiltration.

Sun et al.^[166] have pioneeringly integrated phototherapy with an artificial antigen-presenting system, propelling therapeutic strategies to actively stimulate immune responses. The developed saDC@Fs-NPs constructs super artificial DCs utilizing genetically engineered tumor cell membranes. These NPs not only induce ICD through PDT to release tumor antigens but also directly activate T cells via surface-expressed pMHC-I/CD86 molecules. Furthermore, they synergize with an anti-LAG3 antibody to alleviate immune checkpoint inhibition (Figure 10A). In the 4T1 breast cancer model, this design achieved a 3-fold increase in CD8⁺ T-cell infiltration and activated the abscopal effect, indicating a transition from simple immune priming by phototherapy to active immune regulation. Subsequently, Zhang et al.^[167] developed TPE-Ni/OPYO NPs, innovatively combining PTT with chemoimmunotherapy. This system precisely releases chemotherapeutic drugs through a 940 nm laser-triggered phase transition at 64 °C, employing the ICD effect induced by PTT to promote DC maturation to 89.9%. It achieves spatiotemporal synergy between physical ablation and chemoimmunotherapy, enhancing the abscopal tumor inhibition rate to 68% (Figure 10B).

Advancements in genetic engineering technologies have propelled the field of photoimmunotherapy, enabling a transformative progression from targeted delivery to multidimensional immune programming. Cui et al.^[168] employed metabolic glycoengineering to genetically modify tumor cells, facilitating the overexpression of azide groups on their surface. Based on this study, they developed NP-DBD NPs carrying dibenzocyclooctyne (DBCO) ligands, achieving tumor-specific targeting through bioorthogonal click chemistry. Surface ligands optimized by clustered regularly interspaced short palindromic repeats (CRISPR) screening further elevated the NP enrichment rate in pancreatic cancer to 82%. Upon activation by an 808 nm laser, these NPs synchronously generate ROS with a quantum yield of 0.89 and induce double-strand breaks in both mitochondrial and nuclear DNA through the AIE effect. This innovation activates the cGAS-STING pathway, increasing the secretion of type I interferons compared with traditional PSs and laying the foundation for precise coupling between phototherapy and innate immunity (Figure 10C). Building on this, Cao et al.^[169] employed lentivirus-mediated genetic recombination technology to transform DCs into bioreactors that continuously secrete DEV-AIE biomimetic exosomes carrying CRISPR interference elements targeting the SOX2 gene. These exosomes, genetically engineered to display CD80/CD86-PD-L1 fusion proteins on their surface, can directly activate T-cell receptors. Concurrently, the internally loaded AIE PS specifically accumulates in the tumor stem cell microenvironment, reducing the proportion of CD44⁺/CD24⁺ cancer stem cells through photodynamic effects and achieving spatiotemporal synchronization between innate immune signaling and artificial phototherapy regulation (Figure 10D). Furthermore, Sun et al.^[170] integrated multiple genetic editing strategies to develop the SEx@Fc-NPs engineered vesicle system. This system blocks the CD24 immune checkpoint using Siglec-10 nanobodies delivered by AAV vectors and continuously activates the gp96 molecule using CRISPRa technology to enhance antigen presentation efficiency. Combined with the PDT-induced conversion of M2 macrophages to the M1 phenotype, this system employs a 3D synergistic mechanism of immune remodeling-antigen release-checkpoint blockade. In a melanoma model, it maintained the proportion of central memory T cells above 35% for 42 days, establishing long-lasting cross-tumor immunologic memory with a single treatment. These 3 studies, through layered advancements in genetic editing technologies—ranging from targeted activation and cellular factory modification to the construction of systemic immune networks—collectively propel photoimmunotherapy into a new era of programmable, multitarget regulation. Recent research is accelerating the development of multimodal intelligent platforms, particularly in the field of photoimmunotherapy. Diverse research groups are contributing to this progress by exploring innovative approaches from multiple angles. For example, Wang et al.^[171]. developed a M@P nanoplatform with pH-responsive lysosome-targeting characteristics, which was mainly formed through the self-assembly of an AIE PS (MTCN-3) and an immune adjuvant-Poly (I:C) M@P disrupts lysosomal function by generating ROS and local hyperthermia (56.6 °C) under 520 nm laser irradiation. Synchronously, it can induce pyroptosis [via the caspase-1/Gasdermin D (GSDMD) pathway] and ferroptosis (through GSH depletion and lipid peroxidation). This process releases DAMPs, including CRT and HMGB1, which activate DCs. By integrating lysosomal targeting, dual modes of cell death induction, and immune adjuvant synergy, this study overcomes the limitations of traditional phototherapy’s single mechanism and marks a transition of photoimmunotherapy toward systematic regulation (Figure 10E). Following this research strategy, Li et al.^[172] developed a CMNPs/PD-1 system, which attained novel breakthroughs in molecular modification and targeting mechanisms. This system innovatively integrates genetically engineered PD-1-modified cancer cell membrane vesicles (PD-1/CMNVs) with AOTTIT, a multimodal phototheranostic molecule with high absorption in the NIR-II window. The AOTTIT molecule, through an expansion of its π-conjugated architecture and strategic alkyl chain modifications, achieves a triad of functionalities: NIR-II fluorescence imaging with an emission maximum at 903 nm, a photothermal conversion efficiency of 64.74%, and type I PDT primarily mediated by hydroxyl radicals. PD-1/CMNVs significantly enhance tumor accumulation through dual mechanisms of homologous targeting and active PD-1/PD-L1 blockade. Building upon the explorations of the preceding teams, Wang et al.^[173] make progress in therapeutic modal integration and immune memory induction. They developed an NIR-II responsive intelligent nanoplatform—MBTP-PA-NPs, which targets and depletes intracellular GSH via thiol-yne covalent reactions, thereby alleviating tumor hypoxia. This multifunctional platform combines type I PDT, PTT, and chemodynamic therapy. It activates the caspase-1/GSDMD-N pathway to induce pyroptosis and depletes GPX4 to trigger ferroptosis, enhancing the release of HMGB1 and ATP through dual cell death mechanisms (Figure 10F). These studies sequentially target critical mechanisms: enhancing immune activation via lysosomal targeting, amplifying immune responses through PD-1 blockade, and sustaining therapeutic effects via metabolic modulation. Collectively, they establish a comprehensive technological framework that spans lesion ablation, immune activation, and immune remodeling, advancing photoimmunotherapy into a new era of multimodal synergy and spatiotemporally precise regulation.

The technological progression in immunotherapy unfolds through 4 distinct stages: initial elucidation of the fundamental mechanisms underlying ICD, subsequent creation of artificial antigen-presenting systems, then advancement to genetically engineered biomimetic carriers, and ultimately culminating in the assembly of multimodal intelligent platforms^[174,175]. Successive stages have surmounted the limitations of prior approaches, retaining phototherapy’s fundamental benefits while advancing therapeutic scope through innovative delivery mechanisms and refined immune modulation. This evolution has culminated in a novel oncological treatment paradigm, harmoniously integrating spatiotemporal precision with systemic therapeutic synergy.